Abstract

This paper presents a new algorithm Artificial Neural Network and Seeded Region Growing (ANN-SRG) to segment cloud patches of different types. This method used Seeded Region Growing (SRG) as segmentation algorithm, and Artificial Neural Network (ANN) Cloud classification as preprocessing algorithm. It can be trained to respond favorably to cloud types of interest, and SRG method is no longer sensitive to the seeds selection and growing rule. To illustrate the performance of this technique, this paper applied it on Chinese first operational geostationary meteorological satellite FengYun-2C (FY-2C) in three infrared channels (IR1, 10.3- 11.3μm; IR2, 11.5-12.5μm and WV 6.3-7.6μm) with 2864 samples collected by meteorologists in June, July, and August in 2007. The result shows that this method can distinguish and segment cloud patches of different types, and improves the traditional SRG algorithm by reducing the uncertainty of seeds extraction and regional growth.

How to Cite
YU LIU, DU-GANG XI, XUE-GONG LIU, CHUN-XIANG SHI, KAI ZHANG, Dr.. Automated Cloud Patch Segmentation of FY-2C Image Using Artificial Neural Network and Seeded Region Growing Method (ANN-SRG). Global Journal of Computer Science and Technology, [S.l.], apr. 2012. ISSN 0975-4172. Available at: <https://computerresearch.org/index.php/computer/article/view/491>. Date accessed: 18 jan. 2021.